ART. V.-Researches on the Platinum metals; by WOLCOTT GIBBS, M.D.

(Continued from vol. xxxiv, p. 342, Nov., 1862.)

THE mass of mixed double chlorids, after the volatilization of the osmium and the separation of the iron and other impurities by washing with a concentrated and cold solution of chlorid of potassium or ammonium, is to be rubbed to fine powder, boiling water added, and the iridium reduced by a dilute solution of nitrite of soda, care being taken not to use more of this salt than is sufficient to convert the iridium salt, IrCl, KCl, into Ir, Cl, 3KCl, and to keep the solution neutral with carbonate of soda. Almost the whole of the chlorplatinate of potassium remains undissolved, while the iridium, rhodium, and ruthenium soluble salts remain in solution. The liquid is to be allowed to cool, filtered, the remaining mass washed with cold water until only the chlorplatinate of potassium remains, and the washings filtered and added to the main solution. If the washings have been carefully performed with small successive quantities of water, very little platinum is dissolved, and the olive-green solution contains chiefly Ir, Cl, 3KCl, Rh, Cl, 3KCl, Ru, Cl, 2KCl, and RuCI, KCl with insignificant quantities of PtCl, KCI.

2

A solution of chlorid of luteocobalt, 6NH,. Co, Cl,, is now to be cautiously added as long as a precipitate is produced: a copious pale buff precipitate is thrown down which settles easily, leaving a yellow or orange-yellow solution containing the luteocobalt salt in small excess. The precipitate is to be washed by decantation, then thrown upon a filter and thoroughly washed with boiling water and afterward with boiling dilute chlorhydric acid. The filtrate and washings are to be evaporated together on a waterbath to dryness. They contain the whole of the ruthenium and platinum present in the original solution. The mass upon the filter, which has a pale buff color, consists of the two insoluble double salts,

2

3

2

6NH3. Co2Cl3+Ir2Cl3, and 6NH,. Co2 Cl2+R12Cl ̧, and is perfectly free from ruthenium and platinum.

This process is based upon the fact that the iridium and rhodium double salts above mentioned are almost absolutely insoluble in boiling water and in boiling dilute chlorhydric acid, while the ruthenium and platinum salts, which have respectively the formulas

6NH. Co2Cl2+3RuCl2, and 6NH ̧. Co2Cl2+3PtCl 2,

are easily soluble.

Palladium also forms with chlorid of luteocobalt a double salt which is easily soluble in dilute chlorhydric acid, and which AM. JOUR. SCI.-SECOND SERIES, VOL. XXXVII, No. 109.-JAN., 1864.

crystallizes from the solution, on cooling, in beautiful orange-yellow granular crystals. The formula of this salt is 6NH,. Co, Cl,+3PdCl. Any traces of palladium which may have been present in the original mass of double chlorids will therefore be found with the ruthenium and platinum salts. When the mixed chlorids have been thoroughly washed, palladium is never present. The sesquichlorid of ruthenium gives no precipitate with solutions of chlorid of luteocobalt, and appears not to form a double salt with the chlorid of this radical, possibly in consequence of the triacid character of luteocobalt and the bibasic character of the sesquichlorid of ruthenium, the potassium double salt being Ru, Cl,+2KCI. All the sesquichlorid of ruthenium present in the mass of mixed chlorids in combination. with chlorid of potassium will therefore be found in the filtrate from the insoluble iridium and rhodium double salts.

The mass of double salts of iridium and rhodium with luteocobalt, after complete washing, is to be brought into a flask and boiled with a strong solution of caustic potash until ammonia ceases to be given off. With a concentrated solution this may be effected in a short time. On addition of an excess of chlorhydric acid the black powder readily dissolves, giving a solution which contains the double chlorids of iridium and potassium and rhodium and potassium, Ir, Cl, 3KCl, and Rh, Cl, 3KCl, together with chlorid of cobalt. The solution is to be evaporated to dryness and the chlorid of cobalt dissolved out by boiling with absolute alcohol. The iridium and rhodium are then to be separated by nitrite of soda and sulphid of sodium or ammonium in the manner already pointed out. The sulphid of rhodium after washing and continued ignition gives pure metallic rhodium.

3

3

The filtrate from the iridium and rhodium salts contains a comparatively large quantity of ruthenium in the form of Ru, Cl, 2KC1 and RuCl, KCl, together with a trace of the ruthenium and platinum double salts, 6NH,. Co, Cl,+3RuCl, and 6NH,. Co, Cl,+3PtCl, The solution is to be evaporated nearly to dryness, boiled with a strong solution of caustic potash, and then treated with an excess of chlorhydric acid, which gives the double chlorids RuCl, KCl, PtCl, KCl and Ru, Cl, 2KC1, together with an excess of chlorid of potassium and a little chlorid of cobalt. This last may easily be removed by alcohol after evaporating the mixed chlorids to dryness. Platinum and ruthenium may then be separated by boiling with nitrite of potash, evaporating to dryness, boiling with dilute chlorhydric acid so as to convert the whole of the ruthenium into RuCl, KCl, neutralizing with carbonate of potash, again boiling with nitrite of potash, evaporating to dryness and dissolving out the double nitrite of ruthenium and potash by absolute alcohol.

The nitrite of ruthenium and potash may then be treated in the manner already described and the ruthenium brought into the form of the double salt of mercury and ruthendiamin, from which the pure metal is easily obtained. This method of separating the platinum metals gives excellent results, but is not free from objection. In the first place it will be remarked that it does not dispense with the employment of the alkaline nitrites, although to some extent it facilitates their use. But the chief objection is found in the necessity of employing very large quantities of chlorid of luteocobalt, a salt which is not to be had in commerce and which must therefore be specially prepared for the occasion.

2

3

This difficulty may be in a great measure avoided by employing the chlorid of luteocobalt chiefly as an agent for the separation of rhodium from platinum and ruthenium, which may be accomplished in the following manner. The mass of mixed double chlorids, after the removal of the iron, &c., by washing, is to be rubbed to a fine powder in an unglazed porcelain mortar and then washed with cold water in small portions at a time until the washings give no sensible reaction for ruthenium, when tested in the manner already described with nitrite of potash and colorless sulphid of ammonium. The washings contain all the ruthenium as RuCl, KCl and Ru, Cl, 2KCl, and all the rhodium as Rh, C1,3KCl, together with a not inconsiderable portion of iridium as IrCl, KCl, and a much smaller quantity of platinum. as PtCl, KCl. The iridium in this solution is to be reduced to sesquichlorid by the addition of a dilute solution of nitrite of soda with a little carbonate of soda to keep the solution as nearly neutral as possible. A solution of chlorid of luteocobalt is then to be added as long as a precipitate is produced, when the whole is to be filtered and the precipitate thoroughly washed, first with boiling water and afterward with water containing a little chlorhydric acid. The precipitate on the filter consists chiefly of the rhodium salt 6NH,. Co, Cl,+Rh, Cl,, with a smaller proportion of the corresponding iridium salt. The mixed rhodium and iridium salts are then to be boiled with a solution of caustic potash as long as ammonia is evolved, treated with excess of chlorhydric acid, evaporated to dryness, the chlorid of cobalt dissolved out by alcohol, and the iridium and rhodium separated by nitrite of soda and sulphid of ammonium in the manner already pointed

out.

3

The filtrate from the insoluble rhodium and iridium salts contains the ruthenium as RuCl, KCl and Ru, Cl, 2KCl, together usually with a small quantity of the double salt 6NH,. Co, Cl, +3RuCl, and of PtCI, KCl. The platinum and ruthenium are then to be separated with nitrite of potash and alcohol by the process already described. This method of employing the

chlorid of luteocobalt is extremely convenient when it is desired to obtain pure ruthenium or rhodium at once from the osmiumiridium.

3

3

The presence of a portion of the ruthenium in the form of Ru, Cl, 2KCl in no wise modifies the application of the above process, because this salt gives no double decomposition with a solution of chlorid of luteocobalt. As the nitrite of soda employed to reduce the IrCl, to Ir, Cl, may exercise a reducing action on the Ru, Cl,, it will be found advantageous after washing out the RuCl, KCl and Ru, Cl, 2KCl, to convert the Ru, CI, 2KCl entirely into RuCl, KCl. This may easily be accomplished by adding a solution of caustic potash in excess and then passing a current of chlorine gas into the liquid until the odor of hyper-ruthenic acid is observed. By adding nitric acid in excess so as to dissolve the black precipitate at first produced, and then evaporating to dryness with an excess of chlorhydric acid, the whole of the ruthenium will be brought into the form of RuCl, KCl.

3

When a solution of chlorid of luteocobalt is added to one containing bichlorid of iridium, an insoluble buff-colored precipitate is thrown down, 'consisting of a salt which has the formula 6NH,. Co, Cl,+3IrCl,. When this salt is digested or boiled with an alkaline nitrite, the bichlorid of iridium is reduced to sesquichlorid, and the salt 6NH,. Co, Cl, + Ir, Cl, is formed, well characterized by its extreme insolubility. In the presence of a large excess of platinum in the form of PtCl, KCl, it is very difficult to reduce iridium completely from bichlorid to sesquichlorid, and even in the presence of an alkaline nitrite the chlorplatinate of potassium, after repeated crystallization, obstinately retains a reddish or deep orange tint arising from traces of the corresponding iridium salt. The presence of the smallest trace of iridium may be easily detected in the platinum salt by dissolving the whole in boiling water and adding a solution of chlorplatinate of luteocobalt, 6NH,. Co, Cl,+3PtCl2, which precipitates only its equivalent of the corresponding iridium salt. The precipitate is to be filtered off and digested with a hot solution of nitrite of soda or potash, a small excess of a solution of chlorid of luteocobalt added, and the double chlorid 6NH,.Co, Cl,+Ir, Cl, thoroughly washed. This process affords a perfectly satisfactory method of separating iridium quantitatively from platinum, and for analytical purposes is more convenient than the separation by an alkaline nitrite and sulphid already described. The quantitative separation of iridium from ruthenium and palladium is also readily effected by the chlorid of luteocobalt, as well as the separation of rhodium from platinum, ruthenium and palladium. I shall return to this subject in treating of the metals of this group separately and will then point out

2

another method of using the chlorid of luteocobalt, which is also deserving of attention.

The separation of the metals contained in the mass of sulphids precipitated in the separation of iridium from rhodium, ruthenium and platinum, by the method already pointed out, may be very conveniently effected in the following manner. The mixed sulphids are to be dried, separated from the filter and intimately mixed in a mortar with an equal weight of a mixture of equal parts of carbonate and nitrate of baryta. The filter is to be burned and the ash mixed with the sulphids and baryta salts. The mixture is then to be ignited in a porcelain or earthen crucible for an hour at a full red heat, and the mass, which does not fuse, treated with strong chlorhydric acid, which dissolves the oxyds of rhodium, ruthenium and platinum completely, leaving only sulphate of baryta. The baryta is then to be precipitated by sulphuric acid, an excess of which must be carefully avoided, and then a solution of chlorid of luteocobalt added as long as a precipitate is formed. The double chlorid of rhodium and luteocobalt may then be filtered off and thoroughly washed with boiling water acidulated with chlorhydric acid. By igniting this salt and dissolving the chlorid of cobalt out from the mass, pure metallic rhodium remains. The platinum and ruthenium in the filtrate may then be separated by means of nitrite of potash and alcohol in the manner already described.

This method of treating the sulphids requires only a small quantity of chlorid of luteocobalt, is extremely easy of application and is much shorter than the first method which I have described. Taken in connection with the process for separating iridium by means of nitrite of soda and sulphid of sodium, it furnishes an easy and complete solution of the problem of the qualitative or quantitative separation of the metals of this group, osmium only being determined by the loss.

Cambridge, Nov. 10th, 1863.

(To be continued.)

ART. VI.-Tubularia Not Parthenogenous; by Prof. HENRY JAMES CLARK, of Harvard University, Cambridge, Mass.

It is with no small degree of pleasure that I announce the discovery of the eggs of the Tubularians. During the middle of October I had in my aquarium the three most common species, of this group, on our shores, viz: Tubularia indivisa Lin. (T. Couthouyi Ag.) Thamnocnidia coronata Ag. (Tubularia coronata Abild., Thamnocnidia spectabilis Ag.) Parypha calamaris? (P. crocea Ag., Tubularia calamaris Van Ben.?). In each of these I have traced the development of the egg, from its inception to